All posts tagged: AI Agents
Everything you can write down as a rule belongs in code, where it can be tested, versioned and audited. Everything that needs judgement belongs to the agent. How we draw that line in Nexus OMAS, and why it decides whether an agent survives production.
We collapsed three overlapping concepts in Nexus OMAS into a single one: the project. Here is the design decision, why we made it, and what it unlocks for the AI agents.
Sinapsis, our project management platform, now connects directly to Claude through the Model Context Protocol (MCP). Teams manage their board conversationally, and clients report bugs, request features and track progress — without leaving their AI assistant.